2017 Hardware Thread

solarstarshines said:

Those boards look great especially the MSI Titanium after looking at a reviewer Jaystwocents been watching him for like 4 years but he said something that made me not choose Ryzen and it rings true with other reviewers

Some are having trouble getting stability even with Bios updates so to me it would be better to wait it out a month or so to see what real problems are showing up

Every new platform has it's Demons i wouldn't run out just yet i would wait to see what everyone thinks about it first at least trusted sources that is

Agreed, v 1.0 of anything is usually questionable. This MB of mine had 6 BIOS upgrades, all in the first year but I have also seen some Asus MBs with dozen or more updates.

essenbe said:

Mike, I really don't know, but I would suspect an unstable overclock.

Could be steve, but I pass intels stress tests ...and its only PerformanceTest 9.0 that gives me a no score in one of its many tests that it runs....

essenbe said:

Mike, I really don't know, but I would suspect an unstable overclock.

He could try Catzilla and see how the physics score pans out there.
And Mike try setting your Ring Ratio(Cache ratio) to 300MHz less than your core multiplier.
For example, if you're running CPU Core 48x(4800MHz or 4.8GHz) set Cache ratio to 45x.


Also you could try setting your OC back to the last known good setting, then try increasing your clock speed bit for bit instead.
This will also affect RAM speeds, and Cache too.
This is how I was able to get to 4.925GHz, I slowly went up to reference clock to 100.500 and had to eventually raise my core voltage too.

solarstarshines said:

See this is where it's going to get a little sticky Intel can always work on a code to improve their Skylake Kabbylake Broadwell-E performance it would be worth it to improve the ipc by 5% if possible with a code upgrade but that is dreaming


I have a feeling though Ryzen won't be as hard punching as people think, It will kick butt for sure but there is going to be a bunch of problems in the first few months because you know they only showed off chips that were solid at the time God knows how many chips they had to sort to get a good one

Well yeah, some "Binning " is inevitable but process should stabilize pretty soon. I thought I lucked out with my FX 6350, 4.6 GHz was doing at 1,370V but at 4.8 already needed 1.44v. It's overclocking that usually exposes weaknesses.
FX 8300 series from base one to 9370 only binning makes difference, everything is physically identical. FX 9590 is almost identical chip too. Difference is almost 2GHz on essentially same chip.

Cliff S said:

He could try Catzilla and see how the physics score pans out there.
And Mike try setting your Ring Ratio(Cache ratio) to 300MHz less than your core multiplier.
For example, if you're running CPU Core 48x(4800MHz or 4.8GHz) set Cache ratio to 45x.


Also you could try setting your OC back to the last known good setting, then try increasing your clock speed bit for bit instead.
This will also affect RAM speeds, and Cache too.
This is how I was able to get to 4.925GHz, I slowly went up to reference clock to 100.500 and had to eventually raise my core voltage too.

Downloading Catzilla now Cliff, I am at x47 now, with the Cache ratio @45x

OldMike65 said:

Downloading Catzilla now Cliff, I am at x47 now, with the Cache ratio @45x

Drop it back to 44 then.